moral animals ideals and constraints in moral theory sep 2004

Moral Animals
Ideals and Constraints in Moral Theory
In Moral Animals Catherine Wilson develops a theory of morality
based on two fundamental premises: first that moral progress implies
the evolution of moral ideals involving restraint and sacrifice; second
that human beings are outfitted by nature with selfish motivations,
intentions, and ambitions that place constraints on what morality
can demand of them. Normative claims, she goes on to show, can
be understood as projective hypotheses concerning the conduct of
realistically-described nonideal agents in preferred fictional worlds.
Such claims differ from empirical hypotheses, insofar as they cannot
be verified by observation and experiment. Yet many, though not all,
moral claims are susceptible of confirmation to the extent that they
command the agreement of well-informed inquirers.
With this foundation in place, Wilson turns to a defence of egalitar-
ianism intended to address the objection that the importance of our
non-moral projects, our natural acquisitiveness and partiality, and
our meritocratic commitments render social equality a mere abstract
ideal. Employing the basic notion of a symmetrical division of the co-
operative surplus, she argues that social justice with respect to global
disparities in well-being, and in the condition of women relative to
men, depends on the relinquishment of natural and acquired advan-
tage that is central to the concept of morality.
Moral Animals will spark fresh debates within philosophy and across
the social sciences.
This page intentionally left blank
Moral Animals
Ideals and Constraints
in Moral Theory
CATHERINE WILSON

CLARENDON PRESS Á OXFORD
3
Great Clarendon Street, Oxford ox2 6dp
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide in
Oxford New York
Auckland Bangkok Buenos Aires Cape Town Chennai
Dar es Salaam Delhi Hong Kong Istanbul Karachi Kolkata
Kuala Lumpur Madrid Melbourne Mexico City Mumbai Nairobi
Sa
˜
o Paulo Shanghai Taipei Tokyo Toronto
Oxford is a registered trade mark of Oxford University Press
in the UK and in certain other countries
Published in the United States
by Oxford University Press Inc., New York
ß Catherine Wilson 2004
The moral rights of the author have been asserted
Database right Oxford University Press (maker)
First published 2004
First published in paperback 2007
All rights reserved. No part of this publication may be reproduced,
stored in a retrieval system, or transmitted, in any form or by any means,
without the prior permission in writing of Oxford University Press,
or as expressly permitted by law, or under terms agreed with the appropriate
reprographics rights organization. Enquiries concerning reproduction
outside the scope of the above should be sent to the Rights Department,
Oxford University Press, at the address above
You must not circulate this book in any other binding or cover

and you must impose this same condition on any acquirer
British Library Cataloguing in Publication Data
Data available
Library of Congress Cataloging in Publication Data
Data available
Typeset by SPI Publisher Services, Pondicherry, India
Printed in Great Britain
on acid-free paper by
Biddles Ltd., King’s Lynn, Norfolk
ISBN 978–0–19–926767–5 (Hbk.) 978–0–19–922809–6 (Pbk.)
13579108642
To my maternal grandparents,
Harry Helson (1898–1977) and Lida G. Helson (1900–1979)
in memoriam
This page intentionally left blank
Preface
The aims of this book are first, to furnish a foundation for moral theory that
is independent of any particular set of moral commitments and second, to
defend a particular version of egalitarianism on that foundation. Though
meta-ethics and political philosophy can be and are often treated independ-
ently, there is a reason for offering a content-neutral theory of moral
judgement and moral practice along with a defence of particular normative
claims. The most powerful arguments against egalitarianism in contem-
porary moral theory gain much of their force from the ostensibly non-
normative theories of the place of the self in the world and the allied
accounts of the nature of moral judgement that frame them.
Moral judgements, according to the descriptive theory advanced here,
form a subset of normative judgements. Unlike aesthetic and non-moral
practical judgements regarding what ought to be done, they reflect the
endorsement of advantage-reducing rules on the part of those who assert

them. Moral rules are rules, one might say, for not getting ahead. Morality is
the system laid down to compensate for the wear and tear that is the
unavoidable by-product of our ordinary strivings, through the imposition
of certain sacrifices and deprivations.
This might seem puzzling. Though the fiercer and darker aspects of
morality were emphasized by anthropologists earlier in the century, what
might be termed Freudian pessimism has lost ground to a conception of
morality as a source of human flourishing. The motivation behind this
equation is understandable; the prohibitory taboos of our ancestors are
viewed with scepticism if not dismay, and there can be no doubt that our
opportunities and well-being depend on the moral behaviour of others
towards us, especially their veracity, impartiality, and benevolence. Yet,
the relationship between morality and flourishing is mediate and qualified,
not direct and unqualified. Observing the norms of finance, cookery,
decorating, and intimate relationships helps us to live good human lives,
while meritocratic institutions enable us to parlay our talents and attributes
into wealth and influence. Morality, as Kant speculated, is for something
other than worldly success, though it is not necessarily for the expression of
our rationality or the use of our noumenal wills. Platitudes regarding human
flourishing obscure much of what is interesting and difficult in morality and
in moral theory. How many of us can claim that the breaking of an
agreement, or some show of partiality, or some occlusion of the truth, has
never helped us to carry on with our lives as we wished to, and that moral
considerations have never held us up?
Humans are disposed to invent, observe, and enforce advantage-reducing
rules of varying degrees of stringency, and they hold high status in the
hierarchy of social norms. Yet, as we might expect given their origins, there
is considerable disagreement as to how far agents can reasonably be required
to restrain and suppress the operation of their natural and acquired powers
for their own enjoyment and benefit. In my local culture, for example, we

agree that one may not walk into unlocked houses and make off with
people’s television sets, but we do not agree on whether the manufacturers
of television sets may set their wages at whatever level they find to be most
profitable. We agree that the well-off have some responsibility for the sick,
poorly educated, and demoralized members of the underclass, but we
disagree on how far their needs should cut into our enjoyments. We believe
that persons should enjoy the attentions of one spouse at a time, no matter
how many others they could attract and maintain with money or savoir-
faire, but we disagree over their entitlement to non-marital friendship or
adventure. Different codes enjoin different degrees of advantage renunci-
ation on members of the communities bound by them or individuals who
subscribe to them. Morality is, in this respect, scalar.
First-order moral argument is sometimes aimed at establishing what to do
when advantage-reducing rules conflict in a moral emergency. The obliga-
tion to do all that one can to save a life may conflict with the duty not to
prolong someone’s suffering by employing showy medical expertise. But
first-order arguments are often addressed simply to the question how moral
to be. Moral rules are such that we often feel burdened by them, resist them,
and produce what are often plausible justifications for our non-compliance,
even when there is no emergency and no conflict between competing
obligations. Strict adherence to a principle of veracity can be highly disad-
vantageous to an agent; loyalty to difficult and demanding friends can prove
exhausting; and requests from worthwhile charitable organizations can be
irritating. We are faced, in other words, with the problem of exigency.
When can aesthetic, prudential, or simple hedonic considerations justify
an exemption from an obligation that has been assumed by an agent, or that
is held to be generally binding? Are we really required to act as the
viii preface
famous modern moral theories, such as Kantianism and utilitarianism, say
we must?

The argument that the burdens these theories impose—including the
burden of submitting all one’s proposed courses of action to the test of an
impersonal theory—are too great for human beings as they are constituted
by nature is frequently cited as a defeater of their seemingly exigent require-
ments, particularly with regard to issues of social justice. Weaker aspirations
with respect to socio-economic and sexual equality have been a striking
feature of recent prescriptive moral theory and the reasoning behind this
lowering of demand levels has been set out with formidable intelligence in
books and articles published over the last two decades and is documented
below. No contemporary moral theorist has presented these meta-ethical
and substantive issues with greater force and clarity than Bernard Williams,
who died as this book was undergoing its last set of revisions.
Williams’s meta-ethical scepticism with regard to moral realism and
demonstrable obligations is defended here as well founded. To a large
extent, it is up to each of us how moral we want to be and what sacrifices
we are willing to incur. Moral theorizing is constrained by what we want,
now, not what our ideally rational selves ought to want, and by what we
find it easy and difficult to do. Nevertheless, it is possible to preserve a good
deal more of the revisionary content of the famous modern moral theories
than Williams and other critics believed to be possible.
To meet the sceptic’s objection to the very idea of an obligation that
could be independent of an agent’s motivational state, I offer a modal theory
of moral judgements that bypasses the question whether moral judgements
or prescriptive theories can be true. The assertion that an action in a given
context is obligatory has both representational and conative content. The
representational content of a moral judgement is given by an idealized moral
world. Roughly, to assert that action ACT is obligatory in circumstances c is
to claim that, in a morally good world otherwise like ours, agents all do
ACT in c. A satisfactory theory of morals is a representation of an ideal world
that is, all things considered, with respect to its advantage-reducing behav-

iour, preferable to rival worlds instantiating different behaviour. Though we
have no direct access to ideal worlds, only to our own, imperfect one, moral
judgements are in principle confirmable. Theorists with different prescrip-
tive commitments disagree on what things are like in a good world. They
advance their favoured candidates, projecting paraworlds, fictional worlds
for which both verisimilitude and moral goodness are implicitly claimed,
preface ix
and argue that they are the best. Though the account offered here has
significant conceptual connections with both modern contractualism and
modern consequentialism, it is formally distinct from them both.
One is obliged to do what a theory says, to the extent that it has been
confirmed, but no further. Unconfirmed moral judgements are mere
authored norms, with few or many adherents. The confirmation of a
moral claim requires only the agreement of reasonable, competent, well-
informed judges as to betterness relations between worlds, not agreement as
to what the criteria of betterness are. Analogously, confirmation in physics
requires agreement by competent judges about a physical phenomenon, not
about confirmation theory, a technical branch of philosophy of science or
statistics. This simplification of the prescriptive theorist’s task ought to be
welcome, for it is no easy matter to gain agreement on betterness relations
between paraworlds.
Later chapters discuss the problem of exigency in terms of the subjective
costs to agents of conforming to particular rules or policies. Heavy subject-
ive costs tend to disqualify policies, but counterweight principles tend to
override agent’s concerns about costs to themselves. Prescriptive moralists
avail themselves of the argument from heavy costs to justify lower demand
levels and counterweight principles to justify higher demand levels than an
assumed set point. Acceptable moral rules need not be universal and can be
relativized to particular reference classes. But prescriptive proposals, even if
they arise from within particular cultural settings and reflect the concerns of

creatures known to be partial to themselves and to kith and kin, presuppose
a detached perspective. There is an anonymity requirement on moral
theorizing, a distinct intellectual pursuit with its own methodology that is
different from the activity of merely deciding what one is going to do. The
requirement implies that the endorsement and propagation of norms that
differentially serve the interests of the particular reference class that endorses
and propagates the norm qualifies as ideology, not moral theory proper. For,
in virtue of knowing that we have powerful interests in how things go for us
as individuals, we know about ourselves that we are disposed to look for
compromises between moral formulas of obligation and self- or class-
interest and that we tend to seek exceptions to prima facie obligations in
the form of exemptions and privileges. The anonymity requirement carries
no implication to the effect that agents in our world exist in a state of
empirical equality. Indeed, the chief reason for adopting it is the observation
that they do not.
x preface
With this meta-ethical framework in place, I turn to a discussion of social
equality. The presupposition in force is that conditions of social dominance
in which some members of human societies have worse lives—less access to
resources, more anxiety, less leisure, worse health—are rooted in our
primate heritage and are exacerbated by technological progress. Human
beings are inclined to coerce others and to take advantage of their labour
when they are able to do so and both the descriptive and prescriptive sectors
of moral theory must build on this assumption. My argument is that
morality steps in where nature and the marketplace fail. The existence of
moral practices and motivations, in other words, presupposes, not a condi-
tion of natural equality, as Hobbes imagined, but a condition of natural and
acquired inequality, in which agents possess, temporarily or over the long
term, natural or situational advantages, including superior strength, intelli-
gence, knowledge, beauty, alliances, power, or wealth.

To wear down the intuition that moral agents exist in a state of natural
equality, I employ two characters, A
1
and A
2
, who engage in various
transactions. They are equals in their enjoyment of basic human goods
and states, but one of them is primus inter pares. A
1
and A
2
cooperate for
Hobbesian reasons—because conflict is expensive and they want to increase
their productive capacity—but their decision to cooperate rather than
compete does not make their relationship morally adequate. The initial
moment of cooperation announces the beginning of their moral problems,
as our interdependency has multiplied ours.
Where social dominance once depended on ferocity, charisma, birth, or
alliance, alone or in combination with one another, modern institutions
reward competence at specialized tasks with power. Presumed competence
is associated in modern societies with the differential enjoyment of author-
ity, prestige, and well-being. Some degree of variance in well-being pro-
duced by meritocratic sorting is, I try to show, defensible. Worlds that
reward meritorious performance are better than similar worlds that allocate
surplus resources according to other protocols. Yet existing distributive
systems fall well short of what can be considered just. For, in the first
place, large sectors of humanity do not participate in these meritocratic
systems. Second, while the tendency of modern institutions to understand
merit as specialist competence, rather than as ferocity, charisma, birth, or
alliance, points to the role of moral influences that moderate crude advan-

tage-taking, meritocratic systems can remain undermoralized. The modern
market economy, and the relationships of employer and employee, investor
preface xi
and worker, husband and wife to be found within it, represent the
modification by degrees of the earliest urban societies founded on two
principles: the agricultural, building, and craft labour of large numbers of
slaves of both sexes, and the domestic labour of nearly all women. The
increase in circulating wealth and in the organization of productive power
has a seemingly intrinsic tendency to increase inequality between classes and
nations, and between men and women. It is naive to maintain that observed
high variance with respect to well-being is the product of a carefully
contrived and well-monitored utilitarian plan to improve the status of the
worst-off, and that it is simultaneously the by-product of a well-functioning
merit-reward system. It is simply the condition we have inherited, modi-
fied, and succeeded in partially moralizing.
The last three chapters are concerned with the fair division of the
cooperative surplus and focus on the question how much variance in
well-being is morally tolerable. They are linked with the descriptive ac-
count of the earlier chapters by the premiss that to have a moral concern is
to be willing to accept a reduction of advantage to benefit another, and by
the premiss that theory choice cannot reflect one’s actual situation. A
morally good world, it is argued, exhibits variations in well-being at the
margins—with respect to access to the doubtful and speculative, but not
possession of the known and necessary components of well-being. Statistical
equality of outcomes is further defended as the only plausible test of fair
procedures. The last chapter returns to the sociobiological themes of the
opening to consider male–female relations, including love, as morally
relevant phenomena. The strengths and weaknesses of the argument from
heavy costs, as it has been advanced in recent years against the demand for
greater social equality between men and women, are assessed in a way that I

hope will encourage philosophers and social theorists to investigate more
thoroughly the relationship between the constraints allegedly imposed by
nature in our actual world and our sense of what is morally right.
xii preface
Acknowledgements
Work on this manuscript was supported by the Canadian Social Sciences and
Humanities Research Council. Many colleagues have discussed its contents
with me and offered important criticisms, including Sam Black and Virginia
Held. Scott Anderson, David Braybrooke, Avi Craimer, Edward Halper,
Cynthia Holder, Peter Vranas, and David Zimmerman commented acutely
on individual sections, repairing many deficiencies and errors, as did a
number of anonymous referees. David Donaldson and Mukesh Eswaran
of the Department of Economics at the University of British Columbia, and
Sarah Hrdy of the Department of Anthropology at the University of
California, Davis have contributed generously of their expertise. I am
especially grateful to Husain Sarkar, who read an entire draft, offering
detailed suggestions for improvement, and to my students, Johnna Fisher
and Tim Christie, for proofreading and pointed queries. Mohan Matthen,
always quick to see the shape of a problem, has been a valued interlocutor
throughout the writing.
Permission to reprint previously published material is acknowledged with
thanks from the following sources: Kluwer Publishing for ‘The Role of a
Merit Principle in Distributive Justice’, Journal of Ethics, 7 (2003), 1–38 and
the University of Calgary Press for ‘The Biological Basis and Ideational
Superstructure of Morality’, in Richmond Campbell and Bruce Hunter
(eds.), Naturalized Moral Epistemology, Canadian Journal of Philosophy suppl.
vol.(2000), 211–44.
This page intentionally left blank
Contents
1. Morality as a System of Advantage-Reducing Imperatives 1

1.1. A Platform for Human Morality 2
1.2. The Demarcation Problem 8
1.3. Are Advantage-Reducing Imperatives Natural? 16
1.4. Morality and Hypermorality 27
2. Paraworlds and Confirmation 35
2.1. Moral Belief-Sets and Theory M 36
2.2. Authored and Unauthored Norms 38
2.3. The Form of a Moral Theory 43
2.4. Moral Theories and Paraworlds 48
2.5. How Remote are Paraworlds from the Real World? 56
2.6. Relativism 62
3. Limits on Theory I: Costs to Agents 70
3.1. Exigency in Moral Theory 71
3.2. The First-Person Standpoint 77
3.3. The Argument from Heavy Costs 85
4. Limits on Theory II: Immanent Standpoints 95
4.1. Immanentism 97
4.2. Fast and Slow Paraworlds 101
4.3. The Disqualification Thesis 104
4.4. Opacity and Disqualification 113
4.5. In Defence of Theory 119
5. The Anonymity Requirement and Counterweight
Principles 126
5.1. The Anonymity Requirement 127
5.2. The Partiality Exemption 142
5.3. From Theory to Practice 147
5.4. Counterweights to the Argument from Heavy Costs 151
6. The Division of the Cooperative Surplus 164
6.1. Is Variance a Moral Concern? 166
6.2. Procedural Theories of Justice 174

6.3. Basic and Symmetrical Cooperation 179
6.4. Equality 189
6.5. Justification and Consent 195
6.6. Immanentism and the Argument from Inevitability 202
7. The Role of a Merit Principle in Distributive Justice 212
7.1. Two Distributive Norms 213
7.2. Instrumental Considerations regarding Merit 228
7.3. Detecting and Assessing Merit 236
7.4. Objective Deprivation and Thresholds 242
7.5. Statistical Equality of Outcomes Required 248
8. Moral Equality and ‘Natural’ Subordination 254
8.1. Male–Female Relations in Moral Philosophy 255
8.2. Are Women Objectively Deprived? 257
8.3. Some Favoured Explanations for Female Subordination 266
8.4. The Argument from Heavy Costs 277
8.5. Recursive Effects of Social Judgement 283
8.6. Policies for Equality 288
8.7. Love as a Morally Relevant Phenomenon 293
Bibliography 302
Index 313
xvi contents
1
Morality as a System of
Advantage-Reducing Imperatives
The theory of morals divides into the descriptive theory of moral
phenomena and moral judgement and the prescriptive theory of what
we ought, morally, to do. Before treating of moral judgements in the
abstract and addressing particular questions of right and wrong, it will
be useful to give some attention to moral phenomena, descriptively
considered.

The concept of morality is related to but not coextensive with the
concepts of care and mutual assistance, both of which have their place
in the non-human world. Yet more seems to be required for moral
observance than the occurrence of actions-that-benefit-anot her or
actions-that-benefit-the-collective. The care of the crow or the wolf
for its offspring benefits them, as the warning call of the goose benefits
the flock, without being moral. Ants and bees live in family groups
whose members must interact with one another to reproduce, to
feed, and to defend themselves and their young, but their cooperation
is no more moral than is the symbiosis of tree and vine.
Nevertheless, the altruistic and cooperative behaviour exhibited by
social animals has a precursor relationship to human morality. The
psychological platform that enables an animal to suppress or moderate
its impulses—especially its aggressive and proprietary impulses—is a
necessary underpinning for morality as we understand it. However
sophisticated or complicated by conditions and exceptions our moral
beliefs are, whatever rationale in terms of long-term happiness and
general flourishing we provide for them, and however great the
satisfactions of morally appropriate behaviour may be, moral emo-
tions and practices involve some degree of repression. An appreci-
ation of this fundamental point is important for progress in the
prescriptive sector of the theory of morals as well as in the descriptive
sectors.
1.1. A Platform for Human Morality
Consider the behaviour observed in modern social primates such as
baboons and chimpanzees.1 These animals have distinctive personal-
ities and recognize each other as individuals. They know who their
children are even after they are grown, and they have friendships and
enmities. Their behaviour is characterized by patterns of loyalty,
reciprocity, and revenge for injury or betrayal. The animals form

coalitions and may come to one another’s defence, but they also
refuse at times to assist each other when help would be useful.
They compete with one another, chasing and biting each other,
snatching each other’s food, or displacing one another from desirable
resting places. They also retaliate against such interference and attack
strangers. Both males and females—but principally females—look
after the welfare of infants; there is also occasional infanticide by
males, as well as loss of infants through bad mothering or carelessness.
The animals take an interest in the condition of their own and each
other’s skin and hair. They take turns grooming each other and can
treat each other’s splinters and abscesses with some success.2 Aggres-
sive interactions between males and males and between males and
females are triggered by feeding competition, or represent redirected
aggression towards another animal. Sometimes one animal harasses or
attacks another for no evident reason.3
In these animals, biological flourishing is compatible with and
perhaps depends on a combination of benign neglect, help, especially
where the effort may be repaid in the future, and harm, especially
where successful retaliation is unlikely. They are neither primarily
1 These details are drawn from M. R. A. Chance and Clifford L. Jolly, Social Groups of Apes, Monkeys
and Men; Barbara Smuts, Sex and Friendship in Baboons; and Wolfgang Koehler, The Mentality of Apes.
2 Koehler describes how a chimpanzee removed a splinter from Koehler’s own finger ‘by two very
skillful, but somewhat painful squeezes with his fingernails; he then examined my hand again very
closely, and let it fall, satisfied with his work’. Mentality of Apes, 321–2. Koehler observes further that ‘[if ]
one is on friendly and familiar terms with an ape who has been injured—say by a bite—one can easily
induce the creature to extend the injured limb or surface for inspection, by making the expressive sounds
which indicate sorrow and regret, both among us and among the chimpanzees’. Ibid.
3 Barbara Smuts, Sex and Friendship in Baboons, 90 ff.
2 advantage-reducing imperatives
selfish, nor primarily altruistic, neither exclu sively partial to kin, nor

indifferent to kinship relations. They are all these things, under
different conditions, and different individuals possess these traits and
dispositions in different measures. They react and respond to oppor-
tunities, threats, or changes in circumstances according to their pre-
sent moods and temperaments, the perceived configuration of the
situation, and their past relationships with others. What an animal
does may not be the right solution to its immediate problem from the
Darwinian point of view. The decision to stay and fight rather than
to flee may result in death; the decision to mate now might result in
its having no offspring who survive to maturity. Over the long run,
however, the combination of personality traits and reactive habits,
as these are distributed amongst individuals in an existing species,
is adapted to the most frequent and the most critical situations
they face.
There is little reason to ascribe moral beliefs or moral agency to
animals that behave in this flexible manner. Only to the anthropo-
morphic eye are there paragons and reprobates amongst them. The
animals cooperate—sometimes. Their cooperation is advantageous to
them as individuals and to their kin—usually. They are aware of each
other’s needs, emotions, and intentions—to some extent. And
human observers can easily develop affectionate relationships with
individual animals. Yet there is no reason to call their animal society a
moral one. This is not because the animals do not have language. For
even if their behaviour was accompanied by verbalizations describ-
ing, sincerely or insincerely, their actions and intentions, this would
not indicate that they had placed themselves under the particular
restraints of morality. Nor is its absence explained by the animals’
inability to ascribe mental states to others.
Missing from their orientation towards the social world is an
interest in regulation as such. There is a certain kind of thought

about themselves that the animals do not have, the thought that social
interactions require the inhibition of spontaneous impulses, whether
these impulses involve aggression or assistance. They may seek on
occasion to control the social behaviour of others, breaking up fights
or engaging in jealous interventions, and they may suppress their own
reactions at times, but it cannot be said of them that they regard the
whole field of social interactions as suscep tible of moulding and
advantage-reducing imperatives 3
determination by themselves as agents. Analogously, it might be said
that chimpanzees do not have aesthetic beliefs or engage in aesthetic
practices, even if they draw or paint when given materials and
opportunity, or decorate their bodies by draping them with ropes
and branches. For they do not see the surfaces of the world—walls,
containers, expanses of skin—as objectionably bare and as calling for
remedial action.
When Hume4 traces the origins of morality to a natural disposition
to perform just and benevolent actions, to approve just and benevo-
lent actions in others, and to attribute merit to those who perform
them, he expresses the view that morality is not exemplified simply in
the performance of actions that happen to benefit others, but requires
a social system that regards actions as items for judgement and
criticism. A Humean might nevertheless protest against the claim
that morality presupposes reflective awareness of social interaction
as a field requiring control of natural tendencies by arguing as follows:
Suppose we were to happen on a group of social creatures somewhat
like humans who possessed speech and reason. Relationships be-
tween members of the group appeared to be friendly and affectionate,
characterized by mutual assistance and devoid of the conflict, physical
aggression, and psychological provocation for which primate soci-
eties, including human societies, are noted.

Suppose these creatures were articulate and explained to us that
their benevolent actions flowed from their sympathetic identification
with the needs of others. Would we not recognize this society as a
moral one, even if its members were not conscious of any struggle to
regulate their behaviour and that of others? The Kantian position is
that there is no morality in this culture, in so far as its members act
from inclination, not from a sense of duty. Nor would their acting
from a sense of duty render them moral, according to Kant, if
dutifulness were simply a special moral emotion unrelated to
thoughts expressible as universal imperatives. As a culture might
lack painting or theatre, and yet be attractive for other reasons—the
extensiveness of its mathematical thought or its melodious songs—
the one described lacks morality and is appealing for other reasons.
Whatever the formal and substantive weaknesses of Kant’s moral
4 David Hume, A Treatise of Human Nature, II. iii. ii. 500.
4 advantage-reducing imperatives
theory may be—and I defer their consideration for later—his position
captures the sense in which morality is an imposition that is not only
an outgrowth and an expression of our natural dispositions but exists
as a corrective to them.
Humans walk upright, talk, laugh, share food, care for their off-
spring for many years, and use their hands for constructive purposes
including building, writing, drawing, and calculating. Their fondness
for normative rules—for doing thing s in the right way, often in
exactly the right way—is manifest in all their activities.5 Whether
we are aware of them and can articulate them or not, our behaviour
and our productions are constrained by internalized canons of appro-
priateness, decency, taste, and civility that forbid us certain actions
that we could easily perform and that deem worthless certain prod-
ucts that we could easily fashion and display. Normative statements

concerning what is fitting, good, meet, appropriate, and right to do
are asserted, inculcated, followed, and enforced, and they are also
scorned, ignored, contested, and evaded. Norms may be explicit and
general; into this joint category fall the international codes of conduct
pertaining to war, the actions by the commanders of ships on the high
seas, and the agricultural regulations of large countries. Or they may
be tacit and restricted, like the telephone protocols followed by a
group of small-town teenagers or the haircut norms of a group of
businessmen. They may be explicit and restricted or tacit and general;
there are norms establishing what it is fitting to eat at different times
of the day and on different holidays, what we talk about and what
words we use, how we greet people, and how we manoeuvre our
bodies through the world. We scan for infractions of the rules of
fittingness and goodness, comment upon them, and punish them,
even if the punishment is only adverse criticism and the rule-breaker
is unaware that he is a subject of critical gossip.
The distinction between a species-specific behavioural regularity
and a widely followed norm is imprecise. The habit of eating within
an hour of arising in the morning and eating again at midday is partly
a physiological requirement for active diurnal creatures, partly a
convention. Exclusive pairings between males and females are natural
5 Allan Gibbard refers in this connection to our ‘broad propensities to accept norms, engage in
normative discussion, and to act, believe, and feel in ways that are somewhat guided by the norms one
has accepted’. Wise Choices, Apt Feelings, 27.
advantage-reducing imperatives 5
for humans, but marriage is a norm that adds extra rigidity to the
typical pattern. Deviance from a species-specific behavioural pattern
tends, however, to reduce the biological fitness of a creature, through
the working of what Bentham termed a natural sanction, while
intentional or unintentional deviance from a norm may actually

enhance biological fitness but has the potential to call forth a social
sanction.6 Overeating shortens life and reduces reproductive oppor-
tunities, but bigamy might well increase both, though in many
countries it is punished by law. The rules humans collectively invent
and propound, and to which they try to hold others, extend beyond
what is necessary either for biological survival or for the persistence
and flourishing of communities. If wealthy businesspersons in
Canada, unlike Italian aristocrats of a former era, eschew the wearing
of ruby pendants, this is not because the practice is biologically
dysfunctional or intrinsically disruptive.
The liking for norms and the pleasure taken in moulding thinking
and acting so that it operates within constraints is evident in the great
human interest in games, in which we take part cheerfully despite
what is often a virtual certainty of losing. Economists like to present
us as chiefly motivated by the desire to obtain preferred goods
through the acquisition of exchangeable currency, but no rational
person would accept the offer of a pile of gold on condition that he
abstain from all normatively structured activity. Even those who enter
lotteries in the vain hope of obtaining a pile of gold seem to take their
chief pleasure in picking their numbers according to some system.
Cognitively, we are equipped to follow rules, and affectively we are
equipped to enjoy following them, and it is not fanciful to think that
the ability to master phon ological and grammatical systems is some-
how connected with a broader facility with rules. Young animals
play, and perhaps they use rudimentary rules or could be taught to use
them, as some chimpanzees can be taught, with effort, to use sign
language. Human children have a broader aptitude for learning new
routines and seem to enjoy the constrained behaviour involved in
dancing, singing, and drawing, as well as in talking. They grow up
into such norm-governed activities as proving theorems, making

6 Jeremy Bentham, Introduction to the Principles of Morals and Legislation, 27 ff.
6 advantage-reducing imperatives
artistic representations, and creating and participating in elaborate
bureaucracies and administrative hierarchies.
Rules encompass regulations, norms, idiosyncratic personal rituals,
and social conventions. Any rule can be asserted in the imperative
voice: ‘You! Do (not) x’.7 The imperative form of any rule is con-
vertible into a declarative form as a value judgement (‘It is wrong
(unseemly, inappropriate, immoral, indecent, incorrect, illicit . . . )
to/not to x’). Not only do humans proclaim and observe rules, they
reflect on them and theorize falsely and truly about their rules. They
make certain assumptions regarding them—for example, how fre-
quently certain rules are likely to be broken—and decisions about
what to do about it when they are. Some rules are known to need
strict enforcement, others are not; some rules are believed to apply
universally, while others are believed to apply only to members of
one community or class.8 The logic of rules is non-monotonic; rules
admit of exceptions, and exceptions to rules admit of exceptions in
turn. Nevertheless, almost all cultures believe that there are some
rules that admit of no exceptions and that bind categorically.9 And
they may give them supernatural or at least supramundane signifi-
cance, insisting, for example, that certain rules were issued by a god,
or are observed by an immaterial substance resident within us, or that
infractions of important rules are automatically lethal for the rule-
breaker, or shameful to his dead ancestors.
Formulas of obligation—statements of the form ‘I (you, he, she, it,
one, we, they) ought to (should, must) do such-and-such’, uttered
aloud, written down in books, implied or precisely articulated in
public discourse—are expressions of social rules and are ubiquitous
in both their hypothetical and their so-called catego rical forms. Rules

stating moral obligations are an interesting and problematic subclass.
There is a greater tendency to regard moral norms and requisites as
issuing from a transcendental source and as commanding universal
human agreement in principle than there is to regard prudential and
aesthetic rules as transcendental or universal. It is often said that moral
7 That moral rules are imperatives backed up by reasons was a major theme of R. M. Hare’s work; see
The Language of Morals, ch. 1 and his retrospective Sorting out Ethics, 12 ff.
8 Robert B. Edgerton, Rules, Exceptions and Social Order, 221 ff.
9 Ibid. 254.
advantage-reducing imperatives 7
rules take precedence over other rules and other considerations. But
which rules are moral rules?
1.2. The Demarcation Problem
The theoretical question what makes a given rule a moral rule—in
virtue of what perceived properties are those who treat it as a moral
rule doing so?—is different from the question whether anyone does
or everyone should endorse the rule. We can agree that ‘Doctors
should not assist their patients to commit suicide’ is a moral rule,
rather than a rule of etiquette, even if we think it is a bad rule or that it
ought to be disregarded under specific conditions. We can agree that
‘Protect your eyes when looking directly at the sun’ is a prudential,
not a moral rule, even if we think that it is a good rule that all sighted
creatures ought to obey. It is difficult, however, to specify the topic
of moral rules, what they seek to regulate, in a way that is non-
committal as between moral theories and that does not import
prescriptive considerations into a descriptive task. Though we
can sort rules into the categories of manners, dress codes, aesthetic
guidelines, professional protocols, game-specific rules, practical in-
junctions, and moral imperatives, it is surprisingly difficult to articu-
late the criteria employed in making such discriminations. The

demarcation problem is not solved by appeal to content. Both
moral rules and taboos are largely concerned with prohibitions
involving sex, killing, and kinship obligations. And certain concep-
tions of virtue or upright living are difficult to distinguish from
specifications of elite manners.
It is sometimes said that moral rules are concerned with how to
behave or how to live, but this specification is vague. To be told
that morality contributes to human flourishing, or upright and decent
living, is not to be informed. All rules—the rules of chess, the rules of
warfare—instruct us about how to behave in various situations, and
both aesthetic and prudential rules (Don’t mix plaids and stripes! Save
your money! Wear a seatbelt!) tell us how to behave and how to live.
There is a wealth of information available from decorators, psycholo-
gists, nutritionists, and government agencies on how to flourish as a
human being. And to be told that morality is concerned with
8 advantage-reducing imperatives